[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Virtual fonts

From: Matej Cepl
Subject: Re: Virtual fonts
Date: Fri, 5 May 2000 22:20:51 +0200

On Wed, 12 Apr 2000 10:00:53 +0200  Tamas Papp wrote: 

> This is the main reason for me. Actually, it's two characters: o
> and u with double acute, and the respective capital letters. There
> are very few _FREE_ latin-2 fonts around, and they are quite easy
> to hack together. I guess that many languages have similar
> problems. 


> Any further comments from interested parties? 

There seems to me three ways, how to deal with ISO 8859-2 
characters in Lout. All are working with unequal results with my 
current setup (RedHat Linux 6.0 CZ, Lout 3.20, Lout2-latin2 from 
jeff's site, Aladdin Ghostscript 6.01, Adobe Acrobat Reader 4.0).  

1) Do nothing. It seems to me, that using @InitialFont { TimesCE 
Base 12p }, lout creates PostScript which I had been able to view 
and print with GhostScript. Unfortunately, it doesn't work well with 
smaller characters (e.g., text in footnotes is screwed up -- probably 
AFM files from Jeff are not well prepared on Latin2) and it doesn't 
work at all with PDF (ps2pdf produces PDF with Type3 fonts which is 
ugly in Acrobat Reader). This does not work with fonts from Aladdin 
GhostScript 5.50 (which I have still in Linux; certainly it doesn't work 
with GS 4.* on my Win 3.11 notebook).  

2) Download my yal2f-0.9.zip fonts (on 
http://www.volny.cz/cepls/ps-pdf/yal2f-0.9.zip ). These are the last 
incarnation of famous Peter Soos' remodellation of Windows 
TrueType fonts, so that they contain all ISO 8859-2 characters and 
some other usefull characters. Included in my package is 
TimesNewRoman, Arial and CourierNew fonts. Concerning 
characters mentioned by you, there is this in TimesNewRoman-
Medium file  

C 213 ; WX 652 ; N Odoubleacute ; B 32 -14 618 792 ;
C 219 ; WX 652 ; N Udoubleacute ; B 5 -14 642 792 ;
C 245 ; WX 452 ; N odoubleacute ; B 31 -13 420 613 ;
C 251 ; WX 452 ; N udoubleacute ; B 1 -13 451 613 ;

I have included a lot (1777) kerning pairs created by a2ac program 
(available on http://math.feld.cvut.cz/olsak -- page has Czech as 
English references to particular resources; if you want some 
introduction to Type1 mysteries check this package as well as 
t1accent by the same author; both have incredibly interesting and 
well written documentation both in Czech as well as in English).  

BTW, if someone on the list would be so kind and checked these 
fonts (it is a first draft and there are still some problems pending -- 
see below), I would be very gratefull. Also, if somebody with more 
adequate tools than my t1utils & EMACS checked whether 
everything is OK (or, better to say, whether there are some other 
problems than indicated below), it would be great help.  

These fonts are not great, however, they seem to be working and I 
have been to created PDF file (via gswin32 6.01) with them. Now, it 
is for me the best usefull solution (works even on Win 3.11 with the 
last 16-bit GhostScript).  

3) I am not sure, whether it would not be possible to persuade 
some fonts distributed with TeX (there are for example cs fonts 
distributed as part of CSTeX, which is package for support of Czech 
in TeX; these fonts are localization of original cm fonts by D. Knuth 
created by change of the Metafont sources). Unfortunately, it seems 
to me that encoding vector of TeX's fonts (based on original one of 
Don Knuth's cs fonts) is so strange, that its change would be 
inadequate (/space missing ?!), especially in view of the last option.  

4) There seems to be huge step forward in URW fonts distributed 
with Aladdin GhostScript 6.01. They seem to contain all Latin-2 
characters. It seems to me, that copying .AFM files distributed with 
GhostScript to lout's fonts/ directory and (possibly) some tweaking 
of .LCM files (I propose the name GSLatin2.LCM), should do the 

Looking through AFM file with following header, I have found the 
characters you mentioned with following names (there are no 
composites defined in AFM file, so it seems to me, that actual should 
be in .pfb files):  

Comment Creation Date: 12/22/1999
FullName Nimbus Roman No9 L Regular
Version 1.05
Notice (URW)++,Copyright 1999 by (URW)++ Design & Development

C -1 ; WX 722 ; N Ohungarumlaut ; B 34 -14 688 890 ;
C -1 ; WX 722 ; N Uhungarumlaut ; B 14 -14 705 890 ;
C -1 ; WX 500 ; N ohungarumlaut ; B 29 -10 470 678 ;
C -1 ; WX 500 ; N uhungarumlaut ; B 9 -10 479 678 ;

However, I have not tested this method (having not installed new 
fonts in Linux yet), but I do not know, why it could not work. Any 
comments from the list?  

            Have a nice day


Concerning my yal2f-0.9 fonts I have got these comments from 
some reader of Czech TeX list (I have no time to deal with them and 
with small exceptions neither clue how to do it).  

- problems with quotes (especially /quoteleft is incorrect)

- missing lcaron in CourierNew-Medium font

- not equal height among individual families of fonts

- glyphs are not named according to WGL4.0 (I have no idea,
  what's that)

- missing /euro (however, I think, that Lout makes Euro character
  programatically, so it doesn't matter, does it?)

- tags Notice, Version and ItalicAngle are not updated

- there are some kerning pairs in AFM fonts for glyphs which are
  not included in pfb fonts (does it matter?)

- some correction done via following sed script

  mkdir -p newdir
  for ii in *.afm; do
's/^Em/CapHeight/;s/^Descent/Descender/;s/^Ascent/Ascender/' \
      $ii > newdir/$ii

  I have no idea, what is CapHeight, Descender or Ascender
- wrong quoteleft (should not it be like half of quotedblleft?)

reply via email to

[Prev in Thread] Current Thread [Next in Thread]